Probably been posted before but this is direct from trial...

Probably been posted before but this is direct from trial registry.

No formal statistical sample size estimation has been performed due to the exploratory nature of this study. Rather, the sample size is based on clinical and practical considerations.

This study is descriptive in nature, and no formal hypothesis testing will be performed.

The descriptive summary for the categorical variables will include counts and percentages. The descriptive summary for the continuous variables will include number of subjects (n), means, medians, standard deviations, and minimum and maximum values. All data will be listed for all subjects.
For the summary statistics, the number of decimal places will be the same as for the source data for minimum and maximum. Mean will be presented with 1 more decimal place than the source data and SD with 2 more decimal places. Percentages will be presented with 1 decimal place.
For all primary, secondary and exploratory analyses, 95% CI will be presented unless stated otherwise.
Change from baseline in biomarkers in synovial fluid at Days 56, will be analysed using a restricted maximum likelihood (REML)-based mixed model for repeated measures (MMRM) approach. Change from baseline in biomarkers in synovial fluid at Days 168, and in biomarkers in serum and in urine at Days 56 and 168 will be analysed using the same REML-based MMRM approach.
For the analysis of change from baseline using WOMAC Osteoarthritis Index NRS (Total index, pain, functional, stiffness subscale) at Days 11, 25, 39, 56, 84, 112, 140, 168, 266, and 365 it will be analysed using a REML-based MMRM approach. For analysis involving improvement and reduction in subscale from Baseline, logistic regression will be used with treatment as factor and baseline score as covariate will be used.
Other associations between biomarkers and outcomes will be explored.
Pearson correlation coefficient will be used to examine the strength of linear association between the variables of interest. Checks on the normality assumption of these variables will be done before using Pearson correlation coefficient. In the case of skewed data, Spearman correlation coefficient will be used instead.